From 5f1eea4d08e8fd3ba9c0653db78f68b11e527c08 Mon Sep 17 00:00:00 2001
From: Carles Fernandez <carles.fernandez@cttc.es>
Date: Mon, 27 Jan 2025 16:53:37 +0100
Subject: [PATCH] Avoid potential division by zero

---
 .../acquisition/gnuradio_blocks/pcps_acquisition.cc         | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/src/algorithms/acquisition/gnuradio_blocks/pcps_acquisition.cc b/src/algorithms/acquisition/gnuradio_blocks/pcps_acquisition.cc
index 569319d1c..231d3ee18 100644
--- a/src/algorithms/acquisition/gnuradio_blocks/pcps_acquisition.cc
+++ b/src/algorithms/acquisition/gnuradio_blocks/pcps_acquisition.cc
@@ -37,6 +37,7 @@
 #include <array>
 #include <cmath>  // for floor, fmod, rint, ceil
 #include <iostream>
+#include <limits>
 #include <map>
 
 #if USE_GLOG_AND_GFLAGS
@@ -529,7 +530,10 @@ float pcps_acquisition::max_to_input_power_statistic(uint32_t& indext, int32_t&
         {
             doppler = static_cast<int32_t>(d_doppler_center_step_two + (static_cast<float>(index_doppler) - static_cast<float>(floor(d_num_doppler_bins_step2 / 2.0))) * d_acq_parameters.doppler_step2);
         }
-
+    if (d_input_power < std::numeric_limits<float>::epsilon())
+        {
+            return 0.0;
+        }
     return grid_maximum / d_input_power;
 }